Health literacy, medication adherence, and quality-of-life scores in Jordanian adults with chronic diseases: a dual analytical study using the Long-Term Conditions Questionnaire and the EuroQol five-dimension three-level questionnaire

Aims: This study examined the association of health literacy with quality-of-life outcomes among Jordanian adults with chronic diseases using two patient-reported outcome measures, the Long-Term Conditions Questionnaire (LTCQ) and the EuroQol five-dimension, three-level questionnaire (EQ-5D-3L). It also compared conventional regression and machine-learning (ML) approaches for modelling these relationships. Materials and methods: In this cross-sectional online study, Arabic-language questionnaires were completed by Jordanian adults aged 18 years or older with at least one chronic physical, mental, or neurological condition. The survey, designed to take 10–12 minutes, included the LTCQ, EQ-5D-3L, Medication Adherence Report Scale (MARS-5), and 12-item Health Literacy Scale (HLS-Q12). A total of 901 complete responses were analyzed. Data were split into training (n = 633) and test (n = 268) sets. Multiple linear regression was used for LTCQ, Tobit regression for EQ-5D-3L index scores, and ML models were trained to explore nonlinear associations. Results: Health literacy and medication adherence were positively associated with both LTCQ (beta = 3.923 and 2.219, respectively; both p < 0.001) and EQ-5D-3L index scores (beta = 0.025, p = 0.001; beta = 0.020, p = 0.009). Number of medications was inversely associated with both outcomes. In 10-fold cross-validation, Extreme Gradient Boosting (XGBoost) showed the lowest average root mean square error (RMSE) for LTCQ (9.18; coefficient of determination [R²] = 0.384) and EQ-5D-3L (0.172; R² = 0.193), although differences between the leading models were modest. Conclusion: LTCQ and EQ-5D-3L showed overlapping but distinct association patterns. LTCQ was more strongly linked with psychosocial and self-management variables, whereas EQ-5D-3L was more closely aligned with medication burden and clinical comorbidity. These findings should be interpreted as associative rather than causal and as differences in captured dimensions rather than evidence of instrument superiority.